EAGER: Enhancing Generative Event Extraction via Reinforcement Learning with Verifiable Rewards

Omar Adjali; Siting Liang; Omair Shahzad Bhatti; Daniel Sonntag

cs.CLSep 24, 2026

EAGER: Enhancing Generative Event Extraction via Reinforcement Learning with Verifiable Rewards

Authors: Omar Adjali, Siting Liang, Omair Shahzad Bhatti, Daniel Sonntag

Organizations: German Research Center for Artificial Intelligence (DFKI), Germany · Carl von Ossietzky Universität Oldenburg, Germany

Abstract

End-to-end event extraction remains challenging for large language models as it requires simultaneous identification of event triggers, classification of event types, and extraction of schema-grounded argument spans. We present EAGER, a reinforcement learning framework for generative event extraction that combines fine-grained verifiable rewards with Schema-Contrastive Advantage Estimation to alleviate advantage collapse under sparse binary rewards. Our reward design explicitly targets structural validity, extraction accuracy, groundedness, coverage, over-generation, and span precision. Experiments across seven benchmark datasets show that EAGER consistently outperforms prompting, supervised fine-tuning, and prior reinforcement learning baselines, achieving a substantial improvement over the strongest prior method. Results demonstrate that task-aligned verifiable rewards and contrastive advantage estimation substantially improve structured extraction.

Figures & tables

Figure 1: Average-F1 performance of frontier LLMs on end-to-end event extraction across 7 datasets and domains.

Figure 2: Structure of a prompt for end-to-end Event Extraction, comprising a task instruction a set of event schemas and the input text.

Figure 3: Cumulative Mean ACR during training

Reward Component	Notation	Formulation	Role in Training
Validity	$R_{\text{fmt}}$	Binary: 1 if output parses into the required structured representation, 0 otherwise	Filters malformed outputs before any extraction scoring
Extraction Accuracy	$R_{\text{EE}}$	Macro-average F1 over trigger identification (TI), trigger classification (TC), argument identification (AI), argument classification (AC), AI + , and AC +	Captures canonical end-to-end extraction quality
Groundedness	$R_{\text{grd}}$	Verifies that predicted triggers and arguments appear verbatim in the source input; averages span-level support scores	Suppresses hallucinated spans and improves factual consistency
Over-generation	$R_{\text{ovr}}$	Penalizes predictions that exceed the gold event and argument counts	Reduces false positives and limits excessive output length
Coverage	$R_{\text{cov}}$	Proportional recall over gold events and arguments, normalized to prevent inflation from over-prediction	Encourages completeness while balancing over-generation penalties
Span Precision	$R_{\text{span}}$	Token-level Jaccard similarity with an additional penalty for oversized predicted spans	Rewards near-correct boundary predictions and discourages span inflation

Table 1: Summary of the proposed verifiable rewards used in our RL framework for generative event extraction. Each reward targets a distinct aspect: structural invalidity, extraction inaccuracy, hallucination, over-generation, under-coverage, and boundary imprecision.

Few-shot Prompting
Model	WikiEvents	PHEE	CASIE	GENIA11	GENIA13	MLEE	M2E2	Avg.
o1 Jaech et al. (2024)	13.08	2.86	5.49	10.27	7.95	12.80	14.24	9.52
GPT-4o Hurst et al. (2024)	10.31	2.00	3.47	12.43	10.87	13.35	11.97	9.20
GPT-5.4-mini Singh et al. (2025)	10.58	18.86	5.38	11.95	10.56	10.70	13.86	11.70
GPT-5.4 Singh et al. (2025)	8.30	1.42	6.22	11.94	8.74	11.90	17.74	9.47
GoLLIE-7B Sainz et al. (2024)	13.65	30.44	7.47	12.58	9.69	8.49	32.14	16.35

Table 2: Mean F1 results for end-to-end event extraction on the test split of the seven benchmark datasets. Avg. denotes the average F1 performance across datasets and EE subtasks. Best results are shown in bold. Green values indicate absolute improvement over the strongest prior baseline. † denotes literature methods we re-implemented.

Setting/Rewards	Wiki	PHEE	CASIE	G11	G13	MLEE	M2E2	Avg
ALL Rewards	21.31	60.36	14.66	26.32	23.39	23.42	46.09	30.79
$R_{\text{EE}}$ + $R_{\text{FMT}}$	20.12	53.66	15.70	26.11	24.10	21.62	30.22	27.36
+ $R_{\text{grd}}$	22.40	57.74	14.94	26.83	22.48	20.02	36.72	28.73
+ $R_{\text{ovr}}$	23.04	58.27	12.96	26.54	22.14	19.33	39.20	28.78
+ $R_{\text{cov}}$	21.38	58.01	16.12	26.34	24.22	21.53	31.52	28.45
+ $R_{\text{span}}$	21.46	58.97	15.91	26.63	24.27	21.40	34.26	28.99

Table 3: Ablation study on the contribution of each proposed individual reward to the global end-to-end event extraction performance. Blue shading denotes improvement relative to the baseline configuration

R_{\text{EE}}+R_{\text{FMT}}

, while amber shading denotes degradation. Darker shading indicates larger absolute change. Bold indicates the best score per dataset. This color scheme logic applies to the subsequent tables.

Setting	Wiki	PHEE	CASIE	G11	G13	MLEE	M2E2	Avg
GoLLIE-7B	13.65	30.44	7.47	12.58	9.69	8.49	32.14	16.35
w/o A.G.	9.14	7.67	3.13	1.64	1.17	2.41	15.54	5.81
EAGER	21.31	60.36	14.66	26.32	23.39	23.42	46.09	30.79
w/o A.G.	19.43	43.66	8.34	17.32	16.96	12.25	40.33	22.61

Table 4: Impact of augmenting event schemes with annotation guidelines.

Dataset	Method	TI	TC	AI	AC	AI+	AC+
WikiEvents	All	38.92	30.83	17.28	14.51	14.32	12.01
	$R_{\text{EE}}+R_{\text{fmt}}$	39.53	32.37	15.33	12.28	11.73	9.47
	+ $R_{\text{span}}$	39.88	35.31	16.96	14.10	12.17	10.37
	+ $R_{\text{ovr}}$	39.22	35.37	18.22	16.12	15.25	14.05
	+ $R_{\text{grd}}$	38.74	34.12	17.92	15.89	14.66	13.09
	+ $R_{\text{cov}}$	41.53	34.65	15.83	13.52	12.20	10.52

Table 5: Detailed results of event trigger and argument performance including TI, TC, AI, AC, AI+, and AC+. Bold indicates the best score within each dataset. Improvement/Degradations are highlighted relative to the baseline configuration

R_{\text{EE}}+R_{\text{fmt}}

Figure 4: Training rewards dynamics.

Setting	Wiki	PHEE	CASIE	G11	G13	MLEE	M2E2	Avg
EAGER	21.31	60.36	14.66	26.32	23.39	23.42	46.09	30.79
w/o SCAE	13.98	50.69	8.67	15.72	15.24	12.22	41.24	22.54
EAGER $R_{\text{EE}}$ + $R_{\text{FMT}}$	20.12	53.66	15.70	26.11	24.10	21.62	30.22	27.36
w/o SCAE	15.49	52.32	9.03	16.44	16.52	11.67	43.46	23.56

Table 6: Ablation study on the contribution of SCAE. Improvement/Degradations are highlighted relative to the corresponding configuration with SCAE.

Reward Added	Avg. $\Delta$ Trig.	Avg. $\Delta$ Arg.
$+R_{\text{grd}}$	$+0.3$	$+1.6$
$+R_{\text{ovr}}$	$-0.4$	$+3.1$
$+R_{\text{cov}}$	$+0.7$	$+0.5$
$+R_{\text{span}}$	$+0.9$	$+4.2$
All	$+2.1$	$+7.4$

Table 7: Average gain over the

R_{\text{EE}}+R_{\text{fmt}}

baseline for trigger-level Avg.

\Delta

Trig. (TI + TC) and argument-level Avg.

\Delta

Arg. (

\sum

AI,AC,AI+,AC+) subtasks, averaged across seven datasets.

Figure 5: Error Change Relative to Baseline ( $R_{EE}+R_{fmt}$ ).

Appendix figures & tables13 assets

Supplementary material from the paper’s appendix.

Appendix

Model	WikiEvents	PHEE	CASIE	GENIA11	GENIA13	MLEE	M2E2	Avg.
GoLLIE-7B w/o AG	9.14	7.67	3.13	1.64	1.17	2.41	15.54	5.81
GoLLIE-13B w/o AG	10.07	7.78	4.00	2.16	0.99	2.88	22.31	7.17
GoLLIE-34B w/o AG	10.46	9.71	4.25	5.26	5.00	4.55	29.37	9.80

Appendix

Table 8: Performance comparison of GoLLIE models without AG.

Figure 8: Analysis of the main metrics for monitoring our RL training dynamics including the mean completion length, the mean reward score, and generation entropy.

Method	Model	GPUs	Batch Size	Epochs	Wall Time (h)
EAGER	GoLLIE-7B	4 $\times$ H100	4	1	$\sim$ 4

Appendix

Table 9: Computational cost of EAGER training on 4 NVIDIA H100 GPUs. Wall time is reported for the full train dataset, which includes the seven dataset training sets.

Dataset	#Docs	#Instances	#Event Types	#Events	#Arguments	Domain
WikiEvents	245	565	50	598	5,501	Wikipedia
CASIE	999	1,375	5	8,469	22,575	Cybersecurity
PHEE	4,827	4,827	2	5,019	25,760	Pharmacovigilance
GENIA2011	960	960	9	13,537	11,865	Biomedical
GENIA2013	20	664	13	6,001	5,660	Biomedical
MLEE	262	286	29	6,575	5,958	Biomedical

Appendix

Table 10: Statistics of the end-to-end event extraction datasets used in our experiments.

Acronym	Error Category	Definition
Event-level Errors
ME	Missing Events	A gold event instance is absent from the model’s output; the trigger and all its associated arguments are undetected.
EE	Extra Events	The model predicts an event instance with no corresponding gold event; a spurious trigger is generated that is not grounded in the annotation.
Argument-level Errors
MA	Missing Arguments	A gold argument span is not predicted for an otherwise correctly identified event; the event is detected but one or more of its argument slots are left unfilled.
EA	Extra Arguments	The model predicts one or more argument spans that have no corresponding gold argument; spurious fillers are generated beyond the gold annotation.

Appendix

Table 11: Definitions of all error categories used in the error analysis (Section 5.3 ). Categories are grouped into five types: event-level errors concerning the detection of event instances, argument-level errors concerning role assignment and completeness, span-level errors concerning boundary precision, faithfulness errors concerning grounding in the source text, and structural errors concerning the syntactic validity of the generated output.

Figure 9: Comparative evaluation of error categories across datasets.

R_{\text{EE}}

R_{\text{fmt}}

Vs. +

R_{\text{span}}

Figure 10: Comparative evaluation of error categories across datasets.

R_{\text{EE}}

R_{\text{fmt}}

Vs. +

R_{\text{ovr}}

Figure 11: Comparative evaluation of error categories across datasets.

R_{\text{EE}}

R_{\text{fmt}}

Vs. +

R_{\text{grd}}

Figure 12: Comparative evaluation of error categories across datasets.

R_{\text{EE}}

R_{\text{fmt}}

Vs. +

R_{\text{cov}}

Figure 13: Comparative evaluation of error categories across datasets.

R_{\text{EE}}

R_{\text{fmt}}

Vs. All

Method	WikiEvents							PHEE							CASIE
	TI	TC	AI	AC	AI+	AC+	Avg.	TI	TC	AI	AC	AI+	AC+	Avg.	TI	TC	AI	AC	AI+	AC+	Avg.
GPT-4o Hurst et al. (2024)	$24.58$	$20.88$	$5.43$	$4.96$	$3.22$	$2.77$	$10.31$	$2.84$	$2.84$	$2.24$	$1.82$	$1.27$	$0.98$	$2.00$	$4.58$	$4.58$	$5.12$	$4.38$	$1.15$	$0.99$	$3.47$
GPT-5.4-mini Singh et al. (2025)	$32.04$	$23.57$	$2.28$	$1.94$	$1.98$	$1.65$	$10.58$	$50.27$	$47.03$	$5.42$	$4.46$	$3.34$	$2.66$	$18.86$	$12.86$	$12.60$	$3.01$	$2.29$	$0.85$	$0.69$	$5.38$
GPT-5.4 Singh et al. (2025)	$26.10$	$19.90$	$0.96$	$0.96$	$0.95$	$0.95$	$8.30$	$4.09$	$3.90$	$0.23$	$0.16$	$0.09$	$0.08$	$1.42$	$13.82$	$13.82$	$3.89$	$3.12$	$1.45$	$1.23$	$6.22$
Gollie-7B Sainz et al. (2024)	$30.17$	$21.32$	$9.96$	$8.62$	$6.57$	$5.28$	$13.65$	$44.15$	$42.48$	$41.73$	$17.74$	$25.30$	$11.21$	$30.44$	$10.07$	$9.40$	$11.01$	$9.23$	$2.71$	$2.42$	$7.47$
Gollie-13B Sainz et al. (2024)	$31.01$	$19.02$	$10.51$	$9.48$	$7.25$	$6.53$	$13.97$	$44.74$	$42.26$	$41.83$	$16.59$	$24.78$	$10.28$	$30.08$	$8.92$	$8.52$	$11.28$	$9.59$	$2.04$	$1.76$	$7.02$

Appendix

Table 12: Full results on WikiEvents, PHEE and CASIE datasets.

Method	Genia2011							Genia2013
	TI	TC	AI	AC	AI+	AC+	Avg.	TI	TC	AI	AC	AI+	AC+	Avg.
GPT-4o Hurst et al. (2024)	$27.67$	$24.55$	$8.23$	$7.35$	$3.52$	$3.28$	$12.43$	$24.92$	$22.01$	$7.53$	$5.98$	$2.46$	$2.32$	$10.87$
GPT-5.4-mini Singh et al. (2025)	$34.85$	$30.24$	$2.38$	$2.01$	$1.15$	$1.06$	$11.95$	$30.07$	$27.01$	$2.06$	$1.43$	$1.61$	$1.20$	$10.56$
GPT-5.4 Singh et al. (2025)	$31.90$	$29.89$	$3.37$	$2.66$	$2.01$	$1.83$	$11.94$	$21.82$	$19.39$	$3.41$	$3.01$	$2.58$	$2.21$	$8.74$
Gollie-7B Sainz et al. (2024)	$27.84$	$19.41$	$11.08$	$8.61$	$4.66$	$3.88$	$12.58$	$20.15$	$14.56$	$10.69$	$6.24$	$4.02$	$2.46$	$9.69$
Gollie-13B Sainz et al. (2024)	$27.44$	$12.70$	$8.74$	$6.71$	$3.92$	$2.44$	$10.33$	$25.20$	$12.14$	$9.01$	$6.13$	$2.91$	$1.99$	$9.56$

Appendix

Table 13: Full results on Genia2011 and Genia2013 datasets.

Method	MLEE							M2E2
	TI	TC	AI	AC	AI+	AC+	Avg.	TI	TC	AI	AC	AI+	AC+	Avg.
GPT-4o Hurst et al. (2024)	$28.02$	$24.49$	$8.79$	$7.57$	$5.99$	$5.23$	$13.35$	$20.78$	$19.91$	$8.67$	$8.56$	$6.97$	$6.91$	$11.97$
GPT-5.4-mini Singh et al. (2025)	$32.60$	$25.50$	$2.56$	$2.55$	$0.49$	$0.49$	$10.70$	$35.16$	$33.59$	$4.26$	$3.58$	$3.61$	$2.97$	$13.86$
GPT-5.4 Singh et al. (2025)	$34.59$	$30.40$	$2.09$	$1.56$	$1.51$	$1.26$	$11.90$	$33.96$	$33.21$	$11.52$	$10.77$	$8.83$	$8.16$	$17.74$
Gollie-7B Sainz et al. (2024)	$25.35$	$13.04$	$5.14$	$3.07$	$2.57$	$1.78$	$8.49$	$51.16$	$47.84$	$27.88$	$24.81$	$21.69$	$19.45$	$32.14$
Gollie-13B Sainz et al. (2024)	$25.86$	$17.99$	$9.01$	$6.61$	$5.41$	$3.90$	$11.46$	$61.63$	$59.82$	$33.56$	$29.69$	$25.98$	$23.05$	$38.96$

Appendix

Table 14: Full results on MLEE and M2E2 datasets.

Explore similar work

Schema-Constrained Document-Level Event Argument Extraction with Lightweight LLM Fine-Tuning
Jul 18, 2026Roberto Pietrantuono, Antonio Guerriero, Pouya SattariCoreference Resolution Large Language Model Fine-Tuning
University of Naples Federico II, Naples, Italy · University of Salerno, Fisciano, Italy
EVENT5Ws: A Large Dataset for Open-Domain Event Extraction from Documents
Apr 23, 2026Praval Sharma, Ashok Samal, Leen-Kiat Soh +1Information Extraction Document
University of Nebraska Omaha, USA · University of Nebraska–Lincoln, USA · The Citadel, USA
LA-RL: Label-Aware Self-Reflection for Reinforcement Learning in Information Extraction
Jul 26, 2026Xiao You, Tianwei Yan, Zixu Shan +2Information Extraction Relation Extraction
Hefei University of Technology, Hefei, China · Chongqing Jiaotong University, Chongqing, China

Jul 18, 2026, cs.CLJ/K move · Enter open · S save

Schema-Constrained Document-Level Event Argument Extraction with Lightweight LLM Fine-Tuning

Roberto Pietrantuono, Antonio Guerriero, Pouya Sattari

University of Naples Federico II, Naples, Italy · University of Salerno, Fisciano, Italy

Coreference Resolution Large Language Model Fine-Tuning Lightweight

Open paper AskPDF arXiv

Event Argument Extraction (EAE) converts documents into structured event records by identifying argument spans and assigning them schema-defined roles. Document-level EAE is challenging due to long-range dependencies between triggers and arguments, cross-sentence context, and strict role constraints, which often lead to boundary errors, uncertainty in roles, and inconsistencies with restricted schemas. In this paper, we study whether mid-sized open LLMs can perform schema-constrained EAE reliably at the document level on MAVEN-ARG. Our approach combines (i) role-set injection in prompts for schema compliance, (ii) parameter-efficient supervised fine-tuning (LoRA) using the same JSON-only interface used at inference, and (iii) deterministic decoding with post-processing that validates JSON, filters invalid roles, de-duplicates arguments, and aligns spans to the document window. Under the official MAVEN-ARG evaluator, fine-tuned mid-sized open models outperform previously reported GPT baselines across mention, entity-coreference, and event-coreference evaluations; our best model (Phi-4, 14B) reaches 42.39% F1 at the event-coreference level. Code to reproduce experiments is publicly available at https://github.com/dessertlab/EAE/.

Abstract

Explore similar work

Schema-Constrained Document-Level Event Argument Extraction with Lightweight LLM Fine-Tuning

EVENT5Ws: A Large Dataset for Open-Domain Event Extraction from Documents

LA-RL: Label-Aware Self-Reflection for Reinforcement Learning in Information Extraction